Access to the PPM Journal and newsletters is FREE for clinicians.
10 Articles in Volume 12, Issue #6
Carpal Tunnel Syndrome
Case Studies in New Daily Persistent Headache
Hormone Testing and Replacement in Pain Patients Made Simple
Management of Prenatal Low Back Pain
Managing the Diabetic Patient with Dementia
Myofascial Pain Syndrome: Uncovering the Root Causes
New Tools for Improving Patient-to-Physician Communication in Clinical Practice
Suicide and Suffering In the Elderly: We Must Do Better
Three Cases Highlight the Challenges Of Treating Rheumatoid Arthritis
Understanding the Sources of Morphine

New Tools for Improving Patient-to-Physician Communication in Clinical Practice

The integration of computerized patient-reported outcomes data into your assessment tools offers a number of potential benefits for the clinical pain practice: promotion of patient-centered care, use as point-of-care screening and monitoring tools, aid for clinical decision-making, means for facilitating communication within multidisciplinary teams, and monitoring quality assurance of patient care.

The integration of computerized patient-reported outcomes data into your assessment tools offers a number of potential benefits for the clinical pain practice: promotion of patient-centered care, use as point-of-care screening and monitoring tools, aid for clinical decision-making, means for facilitating communication within multidisciplinary teams, and monitoring quality assurance of patient care.

The assessment of the functional status of a patient with pain holds the promise of improving patient-to-physician communication and patient-centered care. Unfortunately, the assessment of a patient’s current status and subsequent treatment is a complex task. Pain disorders are often a confluence of medical pathologies, psychiatric symptoms, and psychosocial issues.1-3 The presence of multiple medical conditions, in fact, has been found to significantly affect treatment outcomes and the general functionality of patients.2 For example, chronic pain may have a significant impact on the social, interpersonal, and occupational functioning of patients.

It is well established that all these factors play an integral role in the onset, maintenance, and/or exacerbation of chronic pain.4,5 Naturally, increased understanding of treatment-related factors allows for greater measurement precision of relevant treatment outcomes and related variables. Assessment of health-related outcomes has become increasingly “patient-centered,” with a focus on “health-related quality of life” (HRQL).2,6,7 Self-report measures of health status are increasingly referred to as “patient-reported outcomes” (PROs).8 PROs can take multiple forms in clinical practice: measures of HRQL, health status reports, symptom assessment, patient-reported function or disability, and patient satisfaction.9

While PRO measures have been well accepted in the realm of clinical research, they also have important implications in clinical practice.7,10 The integration of PRO data offers a number of potential benefits for clinical pain practice (eg, promotion of patient-centered care, use as point-of-care screening and monitoring tools, aid for clinical decision-making, means for facilitating communication within multidisciplinary teams, and monitoring quality assurance of patient care).8,9,11,12 This article will review PRO systems and how they may benefit both patients and clinicians.

Table: Beneftis of CAT for patients, healthcare providers, and researchers

Treatment Outcomes Tracking
Systematic tracking of health outcomes has gained increasing traction and attention.13-16 Traditional methods of assessment in pain management practice and other ambulatory care settings include face-to-face interviews and paper-and-pencil questionnaires. In recent years, computerized methodologies have been increasingly used for their abilities “to make assessment results immediately available for the clinical encounter.”9 This automated function is made possible by the “real-time” efficiency with which data are entered, administered, analyzed, and summarized in printout.9

Gwaltney and colleagues have argued that such computerized assessments have several advantages over paper-and-pencil assessments including reduction of missing data and patient selection of multiple answers to an item; ability to simplify more complex skip functions; and the reduction of data-entry burden.8 Multiple studies have also evaluated the measurement equivalence (ie, comparability of the psychometric properties of data) between measures by administration mode.8,17 Specifically, these studies have examined measurement equivalence for a variety of PRO measures and administration-mode comparisons (eg, computer-based [via the Internet at home, within the provider’s office], personal digital assistant, tablet-based, telephone-based or interactive voice recognition–based, and paper-based). In total, the majority of studies have found measurement equivalence by mode of administration.8,17

Computer Adaptive Testing (CAT)
Reeve argues that “we are on the brink of a new era for health outcomes measurement with the availability of CAT-based tools.”18 As an overview of the process of item response theory (IRT)–based CATs, the test begins with an “anchor” or starting item of average “difficulty.” Based upon response, the CAT’s predetermined algorithm will adjust the estimated level of the latent trait for the respondent. The algorithm selects subsequent questions based upon what will best discriminate theta (ie, term for latent trait being measured). This is repeated until a predetermined level of precision is achieved (ie, a termination criterion is reached [90%-95% CI]).19 This sophisticated approach minimizes the number of items administered, which leads to decreased patient burden and administration time associated with testing (Table 1).

Of course, computerized administration carries some concerns about the nature of the patient population being studied (in particular, literacy, reading level, visual ability, familiarity with touch-screen computers, and manual dexterity).9 However, these barriers may be partially overcome by having alternate administration modes available. Outside of these concerns, the development of CATs based upon IRT may provide a means of addressing the limitations of current questionnaires. Through their design, CATs are able to decrease respondent burden while increasing measurement precision.9

Barriers and Benefits
While there are clear benefits to the use of PRO data, significant barriers may exist with its integration into busy clinical practice settings. These barriers involve multiple “logistical complexities” inherent to medical care settings (eg, staff burden; need for immediate scoring, interpretation, and response summarization; patient burden when completing a set of measures; and overall time required for administration through results delivery).8,20 At the same time, there are significant barriers to the implementation of comprehensive outcome tracking and/or screening in outpatient medical care settings, namely time-related concerns (eg, general length of administration, patient burden, interruption/delays in flow of normal care); lack of staff support for scoring; lack of availability of a trained professional for interpretation; and the procedures for results feedback to patients and providers.6,7,21,22

However, developments in both computer technology and statistical methods have made it possible to overcome these barriers.9,10 Several studies have demonstrated the feasibility of using computer-based administration for screening and/or clinical outcome tracking.23-25 Numerous studies also have examined methods that were able to address various identified barriers.8,22 These methods included differing modes of administration, and variations of statistical methods and results delivery. Altogether, a review of this literature suggests that comprehensive assessments of current health status within routine care (in medical or other ambulatory care populations) are much less feasible without the use of efficient systems of administration, collection, and feedback. In part, these potential barriers may be overcome through the use of a computer- or Internet-based assessment platform for outcomes tracking, patient reports, and data aggregation.

Currently, there are multiple computer/Internet-based assessment platforms that can perform these functions. Among these, an existing, free-for-use clinical research initiative developed and maintained through the National Institutes of Health (NIH) is perhaps the most usable/efficient currently available. NIH sponsored a large research initiative, the Patient-Reported Outcomes Measurement Information System (PROMIS), which sought to establish commonality of health outcomes measurement. Specifically, PROMIS developed a set of PROs for use as computer adaptive tests. These PROs were designed for use with all capable patient populations for which their specific measures apply.26

As a primary goal, PROMIS “sought to build item banks that measured key health outcome domains that were relevant and manifested in a wide range of chronic diseases.6,20,26 For a majority of the domains/sub-domains, relevant clinical samples were used to establish the validity and calibration of CAT item banks.26 As a component of the PROMIS initiative, Web-based software was also created “to enable researchers to create study-specific Web sites that could administer PROMIS CATs and other instruments to research participants or clinical samples.”20 The “Patient Report” includes both a graphical summary and displays PROMIS CAT scores based upon patient age- and gender-specific normative samples.20,26 Given the efficient and accessible nature of the PROMIS “Patient Health Report,” it has clear potential as a means of PRO feedback for patients, providers, and/or treatment teams in clinical research studies.

Clinical Application
PRO data can also be framed as measures used in clinical practice.11 PROs “supply valuable information on health status and treatment effects that could not be collected in any other way.”8 First, they measure constructs that are summarily unobservable (eg, levels of pain, fatigue, depressive symptoms) without patient report. Second, in studies that examined the same underlying constructs, PRO measures often had better reliability than many clinician-administered interviews.8 Naturally, such outcome-tracking interventions that positively affect treatment outcomes hold potential benefits for patient care.

Interventions using PRO data and feedback in clinical settings have varied along a number of dimensions; specifically, variation has been in the type of PRO used, who gets feedback, how often clinicians get feedback, whether training for PRO interpretation was conducted, and the nature of the information being fed back.27 Recent literature reviews have sought to establish greater structure for the implementation of PRO-based interventions in clinical practice. A recent review by Greenhalgh of the application of PROs in clinical practice delineated a taxonomy for framing PRO-based outcomes tracking and clinical interventions: level of aggregation of PRO data (eg, individual, group) and whether PRO data feedback is used at the interface between clinician/patient.12 This creates four possible quadrants within Greenhalgh’s proposed taxonomy: individual level of data aggregation using PRO data at clinician/patient interface; group level of data aggregation using PRO data at clinician/patient interface; individual data aggregation not using PRO data at interface; and group data aggregation not using PRO data at interface.12

The first quadrant includes three categories of interventions: screening, monitoring, and promotion of patient-centered care.12 The initiative toward “patient-centered” care is relevant in considering applications of PRO data on the individual level using the clinician/patient interface.12 The 2004 US Policy Initiative emphasized the importance of patient self-management and involvement in care, as well as shared decision-making with healthcare providers. Authors suggested that PRO feedback should be provided to the clinician/patient on multiple occasions correspondent to appointment visits; this practice would create consistency with the intervention and yield more robust outcomes from multiple time points. Additional recommendations included conducting training with healthcare providers for interpretation of PRO data and assessing/adapting the “fit” of PROs for use within a given clinical practice.7,9,12

The second quadrant includes a group-level summary of PRO data aggregation that uses feedback at the clinician/patient interface. This avenue of clinical research uses existing studies to augment clinical decision-making during clinical contact with patients.12 Using a group level of data aggregation as a research intervention has produced a spectrum of results. Specifically, some studies report that such an intervention (ie, clinician using computer-based tools to assist with differential diagnosis, checking drug interactions) leads to better patient outcomes and better response to treatment, improvement in adherence, and a greater level of health engagement by patients. PRO data can also potentially assist treatment providers in their clinical decision-making at the point of care.7,20 Comprehensive tracking of PRO data collected prior to the point of care, and during subsequent appointments, could provide information that is useful for individual patients and/or an overall practice.

The third quadrant refers to PRO data/feedback that is not used during the clinician/patient interface. The primary use of this PRO application is for providing feedback to members of multidisciplinary teams as a means of facilitating communication amongst them. A benefit of this application is that it establishes a common language with which healthcare providers can discuss patient outcomes, concerns with the treatment process, and general goals for care of the patient.12

Finally, the fourth quadrant refers to a group level of data aggregation that is not integrated into the interface between clinicians/patients. The primary applications in this frame are “quality of care” and “effectiveness” studies for individual practices, healthcare consortiums, hospital care, and/or specific treatments. A promising direction within this frame is for the promotion and establishment of a set of common PRO measures for both clinical practice and researchers. Such an initiative would allow for comparison of study findings across randomized clinical trials of the same interventions and across different clinical sites.12

PRO Improves Patient Reporting
It is reasonable to appraise the use of PROs in clinical practice as “communication events.”11 A recent review suggested that “filling out the forms improves patients’ skills at describing their symptoms, such as the skills related to identifying and classifying their symptoms.”11 Furthermore, these authors hypothesized that, in consideration of the healthcare providers’ beliefs, the “use of PROs overcomes the belief that if the patient doesn’t say anything about a symptom, the patient does not think it is a problem.”11 The authors further noted that, in consideration of individual values, using PROs might aid in overcoming values that may make it difficult for men to report their symptoms. The completion of PRO measures “might validate the appropriateness of reporting the symptoms to a physician. For those sub-groups that tend to under-report psychiatric symptoms, completion of an emotional distress–related PRO may, in some regard, fulfill unmet fundamental needs (eg, to have ‘a sense of control over their situations,’ to feel cared for).”11

Improves Patient-Provider Communication
For individuals with chronic pain or disability, clinical research has found that effective communication is critical for the management of pain and fatigue. A recent study qualitatively examined issues related to the manner in which patients and providers communicate about pain and fatigue.28 In particular, this study reported that the inclusion of psychosocial variables in the discussion of biomedical complaints with providers was “strongly valued” by participants. Furthermore, the authors also reported that the majority of study participants felt it would be beneficial for providers to ask about how pain and fatigue disrupted their lives.28

Greenhalgh12 hypothesizes that the occurrence of PRO-related clinician/patient discussion could also lead to increased patient involvement in care decisions and, thereby, patients may experience an increased sense of self-efficacy in their ability to manage their own health.12 Operationally, self-efficacy can be viewed as greater feelings of optimism and beliefs related to perceived ability to control or master an illness threat.

PRO Data Intervention:  Recent Clinical Research
As discussed, multiple factors necessitate targeted outcomes tracking in the chronic pain patient population. In addition to aggregate level benefits, the process of tracking population-specific outcomes—such as psychiatric symptoms, pain-related symptoms, and psychosocial factors—may also benefit pain patients, healthcare providers, and service providers at the point of care. Several authors have extensively reviewed the use of PROs in clinical practice. Evidence from randomized clinical trials of PROs in clinical practice found that their use increased communication and detection of HRQL concerns.12 One randomized controlled trial examined the effect of an ePRO-based intervention in a large sample (n=286) of cancer patients.23 The patients were randomized to an “intervention group” and control group. The intervention involved study patients completing an HRQL measure and other outcomes, including measures of psychiatric symptoms and quality of life specific to cancer populations, on touch-screen computers prior to each appointment at their regular oncology clinic. Summary scores were fed back to the physicians of intervention group patients each time they completed the HRQL and outcome measures. The study investigators found that intervention patients communicated with physicians about questionnaire-related symptoms and pain symptoms with no increase in overall appointment time compared to the control group. Furthermore, results reflected significant improvement in HRQL and emotional functioning for a number of intervention patients. The investigators concluded that “routine repeated HRQL assessments in individual patients is a feasible and effective approach for improving medical practice.”23

Subsequently, Feldman-Stewart and Brundage11 incorporated these findings into their theoretical framework—they argued that the improved emotional functioning might have reflected patients’ greater sense of control over the physicians’ care for patients. Furthermore, this is a significant advantage because improvement in patients’ sense of control over care has been associated with decreased levels of healthcare utilization.11,29

Another recent study found that screening for pain, fatigue, and emotional distress in cancer patients was important for “optimizing management (of care) and reducing the risk of morbidity.”29 In addition, a study of adolescents and young adults in primary care created a computer-administered Behavioral Health Screen that was integrated into an electronic health record.24 This study used an Internet-based platform to administer health risk and psychiatric measures across 13 domains. The investigators used this assessment over multiple time points as a means of screening and tracking progress. Results indicated that the Behavioral Health Screen was valid and able to be practically integrated into its study practice.24

In a more recent study by Swanholm, Noe, and Gatchel,25 it was found that integrating a PRO-based intervention administered prior to the point of care at multiple time points significantly affected adaptive change in several outcome domains for a representative sample of outpatient chronic pain patients. This study incorporated many of the recommendations discussed by Greenhalgh11 and Feldman-Stewart et al12 including computer-based test administration; use of efficient, precise assessment methods (use of PROMIS’ IRT-based CAT, population-specific assessments); use of a structured framework for PRO feedback (two PRO feedback conditions with a chart review–only comparison group, protocol-based feedback, standardized report format); and the collections of outcomes from multiple domains (psychosocial factors, population-specific outcomes, measures of treatment alliance, and performance-based outcomes). For this study, administration, scoring, and storage of patient-reported outcomes were carried out using the Assessment Center from the NIH-sponsored PROMIS research initiative, a secure, Internet-based platform. Administration was computer-based in all phases of this study.25 Significant outcomes included measures of emotional distress and psychological symptoms, pain-related measures of functioning and psychosocial status, treatment alliance (separately reported by patients and providers), and dimensions of health/illness perception. In addition to PROs, significant group effects were observed over time with performance-based measures (walking time change). Observance of the study process and completion time data suggest that the computer-based mode of assessment, type of assessment, scoring, creation of patient reports, and aggregation of data allowed this system of outcomes tracking to be efficiently integrated within a busy clinic environment.25 Findings from this ongoing study highlight the potential patient care benefits that may be associated with this form of treatment outcome tracking and dual feedback intervention.

The integration of efficient technologies and sophisticated assessment tools create the foundation for systematic outcomes tracking. In part, this foundation allows for the remote delivery of innovative follow-up care, real-time analysis of utilization/costs, point-of-care outcome tracking, and PRO data feedback into outpatient pain management practice. As discussed, the integration of PRO data offers a number of potential benefits for clinical pain practice, such as the promotion of patient-centered care, use as point-of-care screening and monitoring tools, aid for clinical decision-making, means for facilitating communication within multidisciplinary teams, and monitoring the quality assurance of patient care.8,9,11,12 Implementation of such a system promotes a patient-centered treatment environment that may be of benefit to individual patients and providers, treatment teams, clinical practice, and service providers.

Last updated on: June 13, 2017
close X